arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2305.07406cs.CRcs.CLcs.LG
收藏

二合一:针对文本生成模型的模型劫持攻击

Two-in-One: A Model Hijacking Attack Against Text Generation Models

CISPA Helmholtz Center for Information Security(亥姆霍兹信息安全中心);Microsoft(微软)

Wai Man Si, Michael Backes, Yang Zhang, Ahmed Salem

AI总结:

本研究提出模型劫持攻击Ditto,将文本分类任务劫持为生成任务,在多个基准数据集上验证其有效性且不损害模型效用。

摘要:

机器学习在从人脸识别到文本生成的各种应用中取得了显著进展。然而,其成功也伴随着各种攻击。最近,一种新攻击被提出,它同时带来了问责性和寄生计算风险,即模型劫持攻击。然而,这种攻击此前仅聚焦于图像分类任务。在本工作中,我们将此攻击的范围扩展到文本生成和分类模型,从而展示其更广泛的适用性。更具体地,我们提出了一种新的模型劫持攻击Ditto,它可以将不同的文本分类任务劫持为多个生成任务,例如语言翻译、文本摘要和语言建模。我们使用一系列文本基准数据集,如SST-2、TweetEval、AGnews、QNLI和IMDB,来评估我们攻击的性能。我们的结果表明,通过使用Ditto,攻击者可以成功劫持文本生成模型,而不损害其效用。

英文摘要:

Machine learning has progressed significantly in various applications ranging from face recognition to text generation. However, its success has been accompanied by different attacks. Recently a new attack has been proposed which raises both accountability and parasitic computing risks, namely the model hijacking attack. Nevertheless, this attack has only focused on image classification tasks. In this work, we broaden the scope of this attack to include text generation and classification models, hence showing its broader applicability. More concretely, we propose a new model hijacking attack, Ditto, that can hijack different text classification tasks into multiple generation ones, e.g., language translation, text summarization, and language modeling. We use a range of text benchmark datasets such as SST-2, TweetEval, AGnews, QNLI, and IMDB to evaluate the performance of our attacks. Our results show that by using Ditto, an adversary can successfully hijack text generation models without jeopardizing their utility.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑